“Rare words” in classical Tamil literature: from the Uriyiyal to the ...

October 13, 2016 | Author: Anonymous | Category: Documents
Share Embed

Short Description

Key words: Classical Tamil, Tolkāppiyam, Uriyiyal, Uriccol, lexicography, polysemy, thesaurus, Tivākaram. 0. Introduct...


Acta Orientalia Academiae Scientiarum Hung. Volume 63 (3), 301 – 317 (2010) DOI: 10.1556/AOrient.63.2010.3.6


This article examines the organisation of the Uriyiyal, which is the 8th chapter inside the 2nd book of the Tolkāppiyam, the most ancient Tamil grammatical work preserved. That chapter, because it provides (approximate) synonyms for 120 “unfrequent words”, many of them polysemic, represents the earliest lexicographical attempt in Tamil. Those 120 words all belong to the category of uriccol, a residual lexical category, examined by the Tolkāppiyam once the nouns (peyarccol), the verbs (viṉaiccol) and the particles (iṭaiccol) have been dealt with in earlier chapters. The final section of the article examines the posterity of the Uriyiyal, which becomes a marginal section in grammatical literature, but finds its full development in the different scholarly landscape which is created with the composition of thesauri such as the Tivākaram and its successors. Key words: Classical Tamil, Tolkāppiyam, Uriyiyal, Uriccol, lexicography, polysemy, thesaurus, Tivākaram.

0. Introduction One of the challenges faced by Tamil Philology is to give an as precise as possible account of the nature of the language which was used in Classical Tamil literature and ∗

This article presents the current state of my thinking concerning one component in the field of Classical Tamil vocabulary studies, on which I have given a presentation at the conference “Letting the Texts Speak” (ELTE, Budapest, 3rd – 5th February 2010). I wish to express here my thanks to the organisers, Csaba Dezső and Mária Négyesi, for inviting me to take part in that fruitful and pleasant conference, and to the Indian Embassy (and especially to Mr Mohan) for making my visit possible. Thanks to them it was possible for me to acquire knowledge from a number of learned colleagues from India and from Europe. I also wish to express my thanks to my colleague and friend Eva Wilden who was not at the conference but who has read the current version and with whom I have been having a long-ongoing discussion on the topics treated here. All errors are of course mine. 0001-6446 / $ 20.00 © 2010 Akadémiai Kiadó, Budapest



which is often referred to as Centamiḻ (see for instance Chevillard 1996, p. 476 and Wilden 2009). At a time when one is sometimes confronted with public statements which characterise Tamil as a language which has, thanks to a miracle, remained until today both “living and classical”, it is probably more conducive to the advancement of learning to stay away from oxymorons, and to concentrate on the differences between Classical Tamil and Modern Tamil, which are in fact two very different languages and cannot be studied fruitfully by using the same methods. Concerning Classical Tamil, one of the possible tasks is to examine how, over the course of many centuries, daily methodical human effort applied to the preservation and transmission of a literary corpus which was probably rooted in a possibly already diglossic natural language with wide dialectal variation, has created a learned (and partly artificial) language, emulating, by its reliance on rote learning, “eternal Sanskrit” and equipped with such tools as grammars (in an extended sense of that term), like the Tolkāppiyam and the Naṉṉūl, and dictionaries (or rather thesauri), like the Tivākaram and the Piṅkalam. More precisely, the human effort examined in this article will be the one attested for the first time in the Uriyiyal, a chapter of the Tolkāppiyam, the most ancient Tamil grammar preserved. That first effort basically consisted in trying to make explicit the meanings (poruḷ) of 120 words belonging to the type called uri-c-col, and further characterised by the author of that treatise as “unfrequent” (payilātavaṟṟai). Those 120 words were explained by making use of 100 other words, considered as “frequent” (payiṉṟavai). Later, probably a few centuries later, we see that effort pursued in the first Tamil thesaurus (or kośa), called Tivākaram, which deals with ca. 9500 words (see Vaiyāpuri Piḷḷai in Madras Tamil Lexicon, Introduction, p. xxx) and will be followed by still more detailed kōśa-s such as the Piṅkalam (dealing with 14,700 words and mentioned in the Naṉṉūl [N459m]).

1. A Short Description of the Uriyiyal The Uriyiyal, which is the 8th chapter inside the TC [Collatikāram, 2nd book of the Tolkāppiyam], starts with a sūtra which announces first that the topic treated will be the items called uric col and then proceeds to characterise those items semantically and morphologically. The designation uric col is not easily translatable and the use of the component uri “acceptable, appropriate, proper” qualifying col “word(s)” might imply, by euphemism or by way of understatement, that those items have a smaller legitimacy1 for being called col than the other items already dealt with in earlier chapters, such as the peyarc col “nouns (lit. name-word)” and the viṉaic col “verbs (lit.


Similary, in the chapter on metrics (Ceyyuḷiyal), inside the 3rd book of the Tolkāppiyam, the items called uriyacai “appropriate acai” are more marginal than the items called iyalacai “natural acai” (see TP314i). The same remark can be made for the restrictive expression āciriya v-uric cīr “cīr which is appropriate for [the] āciriyam [meter]” which contrasts (in TP321i) with the (more normal) iyaṟ cīr. “natural cīr”. Acta Orient. Hung. 63, 2010



action-word)”. Such a hierarchy had already been made apparent by the succession of two sūtras in an earlier chapter of TC, stating that: (1) colleṉap paṭupa peyarē viṉaiyeṉ / ṟāyiraṇ ṭeṉpa vaṟintici ṉōrē (TC158c)2 “The knowledgeable say that [items] fit to be called ‘words’ (col) are those two: ‘name’ (peyar) and ‘action’ (viṉai)”. (2) iṭaiccoṟ kiḷaviyu muriccoṟ kiḷaviyu / mavaṟṟuvaḻi maruṅkiṟ ṟōṉṟu meṉpa (TC159c) “They say that the linguistic items (kiḷavi) [called] iṭaic col and the linguistic items [called] uric col appear (respectively)3 after them and by their side”. Because the peyarc col “nouns”, the viṉaic col “verbs” and the iṭaic col “particles (lit. interstitial words)”4 have already been dealt with in the 5th, 6th and 7th chapters of the TC, the task of the Tamil grammarian is now to give in this 8th chapter a description of the residual component of Tamil vocabulary. That component is, as will appear to anyone who reads the chapter (see Chevillard 1996, pp. 432–469), rather heterogenous. However, the author of the TC manages somehow to give to the general characterisation of the uric cols an appearance of homogeneity, when he explains that: (3) uriccoṟ kiḷavi virikkuṅ kālai // yicaiyiṉuṅ kuṟippiṉum paṇpiṉun tōṉṟip // peyariṉum viṉaiyiṉu meytaṭu māṟi // yorucoṟ palaporuṭ kurimai tōṉṟiṉum // palacol loruporuṭ kurimai tōṉṟiṉum // payilā tavaṟṟaip payiṉṟavai cārttit // tatta marapiṟ ceṉṟunilai maruṅki // ṉeccol lāyiṉum poruḷvēṟu kiḷattal. [TC297c] “When one gives a detailed exposition of the linguistic items (kiḷavi) called uric-col ‘appropriate words’ [one can say]: (A) that they appear (tōṉṟutal) [for expressing] a sound (icai), an idea (kuṟippu) or a quality (paṇpu); (B) that they can be concretely undistinguishable from either noun or verb; (C) that it can happen that one [uric col] is appropriate for several values (poruḷ) or that several [uric col] are appropriate for one [and the same] value; [and finally] (D) that, supporting [the explanation of] unfrequent [uric col] by resorting to frequent ones, one should distinctly explain the values which pertain to each permanent and stable aspect [of an uric col], according to what is specific in each one.” This sūtra, in which we find mentioned in four sections (A, B, C & D) a number of important topics [the semantic “triad” (icai-kuṟippu-paṇpu), the phenomenon

2 The “c” in TC158c point to a numbering done according to the arrangement of the TC (= Tolkāppiyam Collatikāram) in Cēṉāvaraiyar’s commentary. Other abbrevations used here are TE for the Eḻuttatikāram of the Tolkāppiyam and N for the Naṉṉūl, with “i” pointing to a numbering of the sūtras as found in the commentary by Iḷampūraṇar, and “m” to the arrangement of Naṉṉūl in Mayilainātar’s commentary. 3 I adopt the distributive interpretation from Ca. Pālacuntaram (1988, p. 183). 4 The category is divided into 7 sub-categories by TC250c. See Chevillard (1996, pp. 386 – 390) for details.

Acta Orient. Hung. 63, 2010



of (quasi-)synonymy, the question of polysemy, etc.], to which we shall come back later, is followed by another one (TC298c) which states that: (4)

veḷi-p-paṭu col-l-ē kiḷattal vēṇṭā // veḷi-p-paṭa vārā v-uri-c coṉ mēṉa (TC298c) “There is no need of an explanation for the words which are obvious; since [explanations] concern only the words which are not obvious.”

After this, we find a series of 90 sūtras, starting at TC299c and going upto TC388c.5 That enumeration concerns, as alreadys said, 120 uric col, which will be referred to in this article by use of the labels U1 to U120. And we have to make the hypothesis, based on 4, that none of those items could have been considered as having an “obvious” value at the time when this part of the Tolkāppiyam was composed. The exhaustive list, in the Tamil alphabetical order, indicating in each case the citation form seen in the sūtra, and the reference number of the sūtra, is as follows: (5)6 U1: atirvu (TC316c); U2: amartal (TC380c); U3: ari (TC356c); U4: alamaral (TC310c); U5: aḻuṅkal (TC349c, TC350c); U6: āytal (TC330c); U7: icaippu (TC309c); U8: iyampal (TC358c); U9: iyaipu (TC308c); U10: iraṅkal (TC358c, TC359c); U11: ilampāṭu (TC360c); U12: iṉṉal (TC302c); U13: ukappu (TC305c); U14: ucā (TC370c); U15: uyā (TC369c); U16: uru (TC300c); U17: urum (TC365c); U18: uvappu (TC305c); U19: uṟappu (TC347c); U20: uṟu (TC299c); U21: eyyāmai (TC342c); U22: eṟuḻ (TC388c); U23: ē (TC304c); U24: ēṟṟam [VAR: eṟṟam] (TC337c); U25: ai (TC385c); U26: oḻukal (TC317c); U27: oṟkam (TC360c); U28: ōytal (TC330c); U29: kaṭi (TC383c, TC384c); U30: kataḻvu (TC315c); U31: kamam (TC355c); U32: kampalai (TC349c); U33: kaya (TC320c, TC322c); U34: karuvi (TC354c); U35: kali (TC349c); U36: kavarvu (TC362c); U37: kavavu (TC357c); U38: kaḻivu (TC314c); U39: kaḻum (TC351c); U40: kaṟuppu (TC372c, TC373c); U41: kuru (TC301c); U42: kuḻa (TC311c); U43: kūrppu (TC314c); U44: keṭavaral (TC319c); U45: keḻu (TC301c); U46: cāay (TC330c); U47: cāyal (TC325c); U48: cilaittal (TC358c); U49: civappu (TC372c, TC373c); U50: ciṟumai (TC341c); U51: cīrtti (TC312c); U52: cummai (TC349c); U53: cellal (TC302c); U54: ceḻumai (TC352c); U55: cēr (TC363c); U56: ñemirtal (TC361c); U57: taṭa (TC320c, TC321c); U58: tava (TC299c); U59: tā (TC344c); U60: tīrtal (TC318c); U61: tīrttal (TC318c); U62: tuyavu (TC368c); U63: tuvaṉṟu (TC332c); U64: tuvaittal (TC358c); U65: tuṉaivu (TC315c); U66: terumaral (TC310c); U67: tevu (TC345c); U68: tevvu (TC346c); U69: nampu (TC329c); U70: naḷi (TC320c, TC323c); U71: naṉavu (TC376c); U72: naṉi (TC299c); U73: naṉṟu (TC343c); U74: nām (TC365c); U75: niḻattal

5 After this comes a coda of 8 very interesting general sūtras, concluding the chapter, which I unfortunately cannot summarise here. See Chevillard (1996, pp. 463 –469) for a French translation of the sūtras and of Cēṉāvaraiyar’s commentary on them. 6 A majority of those lexical items being polysemic, it would not make sense to provide a translation for them. Their various meanings have to be understood from the simple lexemes enumerated in 8.

Acta Orient. Hung. 63, 2010



(TC330c); U76: nuṇaṅku (TC374c); U77: nuḻaivu (TC374c); U78: nocivu (TC374c); U79: pacappu (TC307c); U80: paṭar (TC340c); U81: paṇai (TC339c); U82: paṇṇai (TC319c); U83: paravu (TC382c); U84: payappu (TC306c); U85: paḻiccu (TC382c); U86: paḻutu (TC324c); U87: pāytal (TC361c); U88: piṇai (TC338c); U89: purai (TC300c); U90: pulampu (TC331c); U91: puṉiṟu (TC375c); U92: pē/pēm (TC365c); U93: pēṇ (TC338c); U94: paiyuḷ (TC341c); U95: poṟpu (TC335c); U96: pōkal (TC317c); U97: matavu (TC377c, TC378c); U98: mallal (TC303c); U99: maḻa (TC311c); U100: mātar (TC328c); U101: mālai (TC313c); U102: murañcal (TC333c); U103: muḻutu (TC326c); U104: muṉaivu (TC386c); U105: mēvu (TC329c); U106: yāṇar (TC379c); U107: yāṇu (TC381c); U108: vampu (TC327c); U109: vaya (TC366c); U110: vayā (TC371c); U111: vaṟitu (TC336c); U112: vārtal (TC317c); U113: vāḷ (TC367c); U114: vitirppu (TC316c); U115: viyal (TC364c); U116: viḻumam (TC353c); U117: viṟappu (TC347c, TC348c); U118: vemmai (TC334c); U119: veṟuppu (TC347c); U120: vai (TC387c).

2. Polysemy and Approximate Synonymy As already said in (4), the strategy recommended by the TC for clarifying the value(s) of an unfrequent uric col is to rely on one or several frequent words. For example, item U24, ēṟṟam, is explained by sūtra TC337c in the following way: (6) ēṟṟa niṉaivun tuṇivu mākum (TC337c) Ēṟṟam is (i.e. ‘means’) either niṉaivu ‘remembering’ or tuṇivu ‘decision’. According to the traditional interpretation, this rather cryptic explanation seems to indicate that the word U24 (ēṟṟam) was considered as being polysemic. The statement made in TC337c appears thus as an illustration of the 4th line of sūtra TC297c.7 Two words are required for explaining the values of U4 and those two words must have been considered as belonging to ordinary basic vocabulary. Similarly, we could illustrate the 5th line of sūtra TC297c8 by making use of TC319c, in which U44 (keṭavaral) and U82 (paṇṇai) are simultaneously explained thanks to the simple word viḷaiyāṭṭu. (7) keṭavaral paṇṇai yāyiraṇṭu viḷaiyāṭṭu (TC319c) Keṭavaral and paṇṇai, those two [are] viḷaiyāṭṭu ‘game’. How this statement must be understood exactly is not completely clear but we seem to have here at least a relationship of approximate synonymy. The statement made in TC319c probably meant that in a poetical utterance containing U44 (keṭavaral),

7 See 3 (first half of C): orucoṟ palaporuṭ kurimai tōṉṟiṉum “it can happen that one [uric col] is appropriate for several values (poruḷ)”. 8 See 3 (second half of C): palacol loruporuṭ kurimai tōṉṟiṉum “it can happen […] that several [uric col] are appropriate for one [and the same] value”.

Acta Orient. Hung. 63, 2010



substituting viḷaiyāṭṭu for keṭavaral would provide an approximately equivalent utterance. If we examine the 90 sūtras in which items U1 to U120 are explained, we can extract from them the following list, in which there are 100 supposedly simple words, P1 to P100, each of them representing one of the possible (approximate) “values” (poruḷ) of one or several “unfrequent” uric col: (8)9 P1: akattiṭutal “embracing” {C: akattīṭu}10 (TC357c); P2: akalam “width” (TC364c, TC376c); P3: accam “fear” (TC365c, TC383c); P4: aravam “noise, clamour” (TC349c); P5: aṟiyāmai “ignorance” {C: aṟiviṉmai} (TC342c); P6: aṟiviṉ tiripu “mental perturbation” {C: aṟivu vēṟupaṭutal} (TC368c); P7: icai “sound” (TC309c, T358c); P8: iṭumpai “distress” (TC353c); P9: iyalpu “nature” (TC313c); P10: irakkam “pity” (TC350c); P11: iḷamai “youth” (TC311c); P12: iṉṉāmai “unpleasantness” (TC302c); P13: īṉṟaṇimai “vicinity of giving birth” (TC375c); P14: uṭku “fright” (TC300c); P15: uyaṅkal “suffering” (TC369c); P16: uyartal “being-high” (TC305c); P17: uyarpu “superiority” (TC300c); P18: uvakai “rejoicing” (TC305c); P19: uḷḷutal “thinking” (TC340c); P20: eñcā poruṭṭu “non-incompleteness” {C: eñcāmai} (TC326c); P21: aimmai “being-thin” (TC356c); P22: aiyam “doubt” (TC384c); P23: oḷi “brightness” (TC367c); P24: karippu “pungency” (TC384c); P25: kaviṉ “extreme beauty” (TC381c); P26: kaḻinta poruṭṭu “pertaining to what no longer exists”11 {C: kaḻivu} (TC359c); P27: kaḷaṉ “place” (TC376c); P28: kātal “loveliness” (TC328c); P29: kāppu “protection” (TC383c); P30: kūrmai “sharpness” (TC383c, TC387c); P31: kēṭu “loss” (TC350c); P32: koḻuppu “fatness” (TC352c); P33: koḷḷutal “taking” (TC345c); P34: kōṭṭam “bending” (TC321c); P35: ciṟattal [uḷḷatu --] “high degree” {C: ciṟappu} (TC314c); P36: ciṟappu “eminence, superiority” (TC353c, TC383c); P37: ciṟitu “slightly” (TC336c); P38: cīrmai “excellence, perfection” (TC353c); P39: cuḻcci “mental deliberation” (TC370c); P40: cuḻaṟci “whirling” (TC310c); P41: celavu “going” (TC340c); P42: ceṟivu “denseness” (TC323c, TC347c); P43: taṉimai “loneliness” (TC331c); P44: tiraṭci “gathering” (TC363c); P45: tuṇivu “decision” (TC337c); P46: tokuti “collection” (TC354c); P47: nacai “affection” (TC329c); P48: naṭukkam “trembling” (TC316c); P49: nilai iṉmai “intempestivity” (TC327c); P50: niṟatturu “colour nuance” {C: niṟa vēṟupāṭu} (TC373c); P51: niṟaṉ “colour” {C: niṟa vēṟupāṭu} (TC307c); P52: niṟaṉ {C: niṟam} (TC301c); P53: niṟaintu iyaṟal “completeness, plenitude”{C: niṟaivu} (TC355c); P54: niṟaivu “abundance, profusion” (TC332c); P55: niṉaivu “remembering” (TC337c); P56: nuṇmai

19 The English translations provided for the 100 items in that list are based on my French translation of Cēṉāvaraiyar’s commentary on the TC. They rely on the reading of a few hundred fragments, quoted by Cēṉāvaraiyar and coming from the Classical Tamil literary corpus or from anonymous sources. See 10, 11, 12, for a sample. (See also Chevillard 1996, pp. 435 – 462.) 10 The additional items provided between braces and marked with a capital C ({C: …}) are those provided by Cēṉāvaraiyar, when the wording of his commentary differs from the wording of the TC. 11 The example given by Cēᐷāvaraiyar for that phrase contains the word “to regret” (iraṅkal).

Acta Orient. Hung. 63, 2010



“thinness” (TC374c); P57: nuṇukkam [uḷḷataṉ --] “slimming” (TC330c); P58: neṭumai “length” (TC317c); P59: nērmai “regularity” (TC317c); P60: nōy “suffering” (TC341c); P61: pakai “enmity” (TC346c); P62: payam-iṉmai “uselessness” (TC324c); P63: payaṉ “benefit” (TC306c); P64: parattal “expanding” (TC361c); P65: piḻaittal “missing (a target)” (TC339c); P66: puṇarcci “uniting” (TC308c); P67: putitu paṭal “appearing-new” (TC379c); P68: putumai “being-new” (TC383c); P69: peṭpu “cherishing” {C: purantarutal} (TC338c); P70: peritu “in a great measure” (TC343c); P71: peruppu “growth, development” {C: peruttal} (TC339c); P72: perumai “greatness” (TC320c); P73: peṟṟu “increase” (TC304c); P74: polivu “lustre, splendour” (TC335c); P75: maṭaṉ “naïvety” {C: maṭam} (TC377c); P76: mayakkam “confusion” (TC351c); P77: miku pukaḻ “panegyric” {C: perum pukaḻ} (TC312c); P78: mikuti “high degree” (TC299c, TC378c, TC383c); P79: mutirvu “being-ripe” (TC333c); P80: muṉṟēṟṟu “engaging-oneself” (TC383c); P81: muṉivu “disgust” (TC386c); P82: meṉmai “softness” (TC322c, TC325c); P83: mēval “desiring” {C: mēvutal} (TC380c); P84: varuttam “distress” (TC344c); P85: varaivu “excluding” (TC383c); P86: vali “strength” (TC344c, TC366c, TC377c, TC388c); P87: vaḻuttu “lauding” {C: vaḻuttutal} (TC382c); P88: valaṉ “fertility” {C: vaḷam} (TC303c, TC352c); P89: vaṟumai “poverty” (TC360c); P90: vaṉappu “beauty” (TC378c); P91: viṭal “abandoning” {C: viṭutal} (TC318c); P92: viyappu “astonishment” (TC385c); P93: viruppu “desirability” (TC362c); P94: viraivu “swiftness” (TC315c, TC383c); P95: viḷakkam “brilliance” (TC383c); P96: viḷaiyāṭṭu “game” (TC319c); P97: vekuḷi “anger” (TC372c); P98: verūup poruṭṭu “being frightened” {C: veruvutal} (TC348c); P99: vēṭkaip perukkam “increase of desire” (TC371c); P100: vēṇṭal “ardent desire” {C: virumputal} (TC334c).

3. The Uriyiyal Seen as a Graph The preliminary facts concerning the Uriyiyal having been given, I shall now try to provide a bird’s eye view of the global information which is provided by it. For that purpose, to the two lists available in 5 and 6, we must add a list of 154 links, each link corresponding to a relationship of approximate synonymy between one U-type item and one P-type item. The structure can be described as a graph having 154 edges and 220 [=120+100] vertices (belonging to two different types) and it is an interesting exercise to try to determine the number of connected components inside that graph. It turns out that the graph contains 69 components, of widely differing sizes: – – – –

42 components containing 1 U-type element connected with 1 P-type element 14 components containing 2 U-type elements both connected with the same P-type element (this corresponds to the case illustrated in 7). 3 components containing 1 U-type element connected with two distinct P-type elements (this corresponds to the case illustrated in 6). 10 bigger components, the biggest of which contains 15 U-type elements, 18 P-type elements and 31 links. Acta Orient. Hung. 63, 2010



The graphic representation given below illustrates some of those types

In this graphic representation, the top element corresponds to example 7 (alias TC319c) already discussed, the small component on the right side corresponds to the statement contained in TC331c, while the bigger component at the bottom (which contains 7 U-type elements, 5 P-type elements and 11 links) condenses the information given by seven distinct sūtras.12

4. Are the Words in the Uriyiyal Really Rare? Before continuing our exploration of the Uriyiyal and of what it can teach us concerning the development of the literary culture of Tamil Nadu, it seems appropriate to examine a question of a different type: “Are the 120 lexemes which we have been examining really rare words?” The quotation mark contained in the title of this article was put there with this question in mind. The question is of course meaningful only if it is made more precise. Some reference corpus or some context of observation must first of all be specified. In the case of the Uriyiyal, the natural corpus seems to be the the Eṭṭut Tokai “Eight Anthologies”, which comprises 2553 poems totalling more than 29,000 lines13 and the Pattup Pāṭṭu “Ten Songs”, which consists of 10 long poems totalling more than 3500 lines.14 A quick estimate shows that the 32,821 lines contained 12

TC320c to TC323c, TC325c, TC347c & TC348c. The detail is: Naṟṟiṇai (400 poems totalling 4173 lines); Kuṟuntokai (401 poems totalling 2498 lines); Aiṅkuṟunūṟu (500 poems totalling 2160 lines); Patiṟṟuppattu (80 poems totalling 1711 lines); Paripāṭal (22 poems totalling 1833 lines); Kalittokai (150 poems totalling 4297 lines); Akanāṉūṟu (400 poems totalling 7156 lines); Puṟanāṉūṟu (400 poems totalling 5441 lines). 14 They are: Tirumurukāṟṟuppaṭai (317 lines); Porunarāṟṟuppaṭai (248 lines); Ciṟupāṇāṟṟuppaṭai (269 lines); Perumpāṇāṟṟuppaṭai (500 lines); Mullaippāṭṭu (103 lines); Maturaikkāñci (782 lines); Neṭunalvāṭai (188 lines); Kuṟiñcippāṭṭu (261 lines); Paṭṭiṉappālai (301 lines); Malaipaṭukaṭām (583 lines). 13

Acta Orient. Hung. 63, 2010



in this ensemble must contain approximately 180,000 word forms.15 Unfortunately, I am not able at the moment to give even a rough estimate of the number of lexemes which this corresponds to.16 However, concerning our list of 120 terms and its place in this corpus, a few (incomplete) remarks can be made. – – –

Some of the words in our list are extremely rare: this is the case for instance with U107 (yāṇu)17 and U102 (murañcal).18 As an extreme case, the term U44 (keṭavaral) is not attested anywhere in the corpus examined here and the commentators to the Tolkāppiyam are able to mention only anonymous quotations. Some of the words are frequent, but not in the meaning which is described in the Uriyiyal. This seems to be the case with U73 (naṉṟu). The form is found more than 150 times in the corpus, but rarely in the usage which the Uriyiyal is pointing to. As Vaiyāpuri Piḷḷai, the editor of the Madras Tamil Lexicon remarked, in the “History of Tamil Lexicography” which is found in the first volume (pp. xxv– liii): “It is interesting to note that some words whose meanings are now very clear are included in this list of ‘hard words, e.g., TĪRTAL, TĪRTTAL, MUḺUTU, PAḺUTU” (Madras Tamil Lexicon, Vol. 1, p. xxv).

We have to admit that trying to decide for each of our 120 terms why it was included in the chapter is not an easy question and that no final global answer will be provided here.

5. Importance of the Semantic Triad “icai, kurippu, paṇpu” After this brief excursion into the realm of lexical statistics, I now come back to an exploration of the original point of view of the TC. Among the 4 facets, labelled A, B, C & D (see section 1, citation 3) which are contained in the characterisation of the items called uriccol, we have already illustrated C (polysemy and quasi-synonymy) 15

The estimate is based on an extrapolation made on the basis of some quick statistics computed from a digital version of the popular “sandhi-split” Rajam edition of the Naṟṟiṇai, in which a few particles have also been separated. Running the Micro-OCP program, a version of the Oxford Concordance Program, we see that the text contains a VOCABULARY of 5842 elements and contains 23,290 WORDS (using the terminology of OCP). There are roughly 5.6 words per line in the Naṟṟiṇai. Since the great majority of the poems in our two lists use the same meter, no great error will be committed in assuming that they have roughly the same number of words per line. 16 Lemmatised digital texts do not seem to be available. And if they were, they would have to be used with a lot of caution, given the still unsatisfactory state of the editions of Classical Tamil texts. Moreover, making decisions concerning the lemmatisation of word forms is an extremely difficult task. 17 The word occurs only once, in Naṟṟiṇai 50-7, as the predicate of a reported utterance “yāṇatu pacalai”. 18 The term occurs only two times and both occurrences are found in the Malaipaṭukaṭām: “murañcu” in line 144 and “murañciya” in line 268. Acta Orient. Hung. 63, 2010



and D (the contrast between the U-type and the P-type of words). Facet B, which concerns their morphology, will be discussed later (see section 7). I am now going to discuss facet A, which concerns the semantic range of the items called uriccol. That semantic range is said by TC297c (cf. supra, 3) to be characterised by the three terms icai “sound”, kuṟippu “idea” and paṇpu “quality” and it is to be noted that the same triad is found, although in a slightly different order, in one of the sūtras of the TE (i.e. Eḻuttatikāram), which is the first book of the Tolkāppiyam, where we have: (9) uyirum puḷḷiyu miṟuti yākik // kuṟippiṉum paṇpiṉu micaiyiṉun tōṉṟi // neṟippaṭa vārāk kuṟaiccoṟ kiḷaviyum // […] // aṉṉavai yellā maruviṉ pāttiya // puṇariya ṉilaiyiṭai yuṇarat tōṉṟā (TE482i) “The linguistic expressions which, ending either with a vowel or a consonant, appearing for the expression of an idea, a quality or a sound, are irregular incomplete words […]19 all those, belonging to the division sanctioned by usage, do not appear intelligibly in the sandhi situation”.20 The main difference between the wordings found in 3 and in 9 lies in the fact that the linguistic items concerned are called uric coṟ kiḷavi in 3 but kuṟaic coṟ kiḷavi in 9. The commentator Iḷampūraṇar thinks that the two must be identified, as shown by the copula (ākiya “which are”) found in the paraphase which he provides (kuṟaiccoṟkaḷ ākiya uriccoṟkaḷ). However, the examples which Iḷampūraṇar provides do not belong to our list of U-type words21 for this type of construction. I shall come back to the discussion of 9 in section 7, while discussing problems of morphology.

6. Cēṉāvaraiyar’s Use of the Semantic Triad Coming back to TC297c and to the sūtras that follow it inside the Uriyiyal, I shall observe that the TC itself does not make use of the 3 labels icai, kuṟippu and paṇpu while explaining each of the 120 U-type items, but that Cēṉāvaraiyar, on the other hand, systematically tells us for each of them in which category it is to be placed. There is a great dissymetry between the 3 labels, because kuṟippu appears in 93 paraphrases, whereas paṇpu appears 13 times and icai only 3 times.22 I shall illustrate the way Cēṉāvaraiyar proceeds by providing a few examples, drawn respectively from the commentaries to TC349c, TC350c and TC352c: (10) «kampalai mūtūr» {{[puṟam:54_1]}} eṉavum, «oru peruñ cummaiyoṭu» {{[ANON]}} eṉavum, «kali koḷ āya(m) malipu tokupu eṭutta» {{[akam:11_4]}} 19 The sūtra deals with 4 specific types of sandhi situations. Since we are interested only in the first one, I skip the three others. 20 This means that no rule will be provided for users of the grammar to understand what to do. The forms have to be known from usage. I discuss the wording of the predicate (puṇariya ṉilaiyiṭai yuṇarat tōṉṟā) in Chevillard (1996, p. 387). 21 They (viṇviṉaittatu, kārkaṟuttatu, ollolittatu) belong the very specific domain of ideophones. See Chevillard (2004, p. 427) for a discussion. 22 The full list is found in Chevillard (2008, pp. 491 –492).

Acta Orient. Hung. 63, 2010



eṉavum, «uyavup puṇarntaṉṟu iv-v- aḻuṅkal ūr ē» {{[naṟṟiṇai:203_11]}} eṉavum, kampalai mutal ākiya nāṉkum aravam ākiya icaip poruṇmai uṇarttum (Cēṉāvaraiyar, @ TC349c)23 “In the examples kampalai mūtūr ‘The ancient town full of clamours’ (Puṟam, 54-1), oru peruñ cummaiyoṭu ‘with a huge acclamation’ (ANON.), kali koḷ āya(m) malipu tokupu eṭutta ‘(like the lamps) held by a sonorous group of [women], happily gathered’ (Akam, 11-4), uyavup puṇarntaṉṟu iv-v- aḻuṅkal ūr ē ‘this noisy village is united to my suffering’ (Naṟṟ. 203-11), the [four words] starting with kampalai express the value of a “sound” (icaip) which is clamour”. (11) «paḻaṅ kaṇōṭṭamu(m) naliya / aḻuṅkiṉaṉ allaṉ ō» {{[akam:66_25/26]}} eṉavum, «kuṇaṉ aḻuṅkak / kuṟṟam uḻai niṉṟu kūṟuñ ciṟiyavarkaṭku» {{[nālaṭi:353]}} eṉavum, aḻuṅkal, aravam ē aṉṟi irakkamuṅ kēṭum ākiya kuṟippum uṇarttum (Cēṉāvaraiyar, @ TC350c)24 “In the examples paḻaṅ kaṇōṭṭamu(m) naliya / aḻuṅkiṉaṉ allaṉ ō (Akam, 6625/26) ‘his ancient benevolence tormenting his [memory], he had regrets, hadn’t he?’, kuṇaṉ aḻuṅkak / kuṟṟam uḻai niṉṟu kūṟuñ ciṟiyavarkaṭku ‘[What is the tongue made of] in the small-[minded] [people] who, while standing next to someone, are capable of neglecting (lit. regretting) the qualities and mentioning [only] the defects’ (Nālaṭi, 353), [the word] aḻuṅkal does not simply express a clamour/wailing (as previously in 10) but also the ideas (kuṟippu) of pitying/lamenting (irakkam) or of loss/omitting (kēṭu). (12) «ceḻum pal kuṉṟam» {{[kuṟun:287_7]}} eṉavum, «ceḻun taṭi tiṉṟa cennāy» {{[ANON]}} eṉavum, ceḻumai, vaḷaṉuṅ koḻuppum ākiya paṇpu uṇarttum (Cēṉāvaraiyar, @ TC352c)25 “In the examples ceḻum pal kuṉṟam (Kuṟun, 287-7) ‘many fertile hills’ and ceḻun taṭi tiṉṟa cennāy ‘red-dog which has eaten a fat piece of meat’, [the word] ceḻumai expresses the qualities (paṇpu) of ‘fertility’ (vaḷaṉ) and of ‘fatness’ (koḻuppu).” These distinctions, and notably the one between kuṟippu and paṇpu, have been clarified in advance by an explanation which Cēṉāvaraiyar had provided while commenting on TC297c. (13) kuṟippu -- maṉattāṟ kuṟitt(u) uṇarappaṭuvatu. paṇpu -- poṟiyāṉ uṇarappaṭuṅ kuṇam (Cēṉāvaraiyar, @ TC297c) “An idea (kuṟippu) is what is aimed at (kuṟittal) by the internal sense (maṉam [skt. manas]). A quality (paṇpu) is the attribute (kuṇam) [of a substance] which is perceived by the instruments (poṟi) [of the senses].”


TC349c: kampalai cummai kaliyē aḻuṅkal / eṉṟivai nāṉk(u) -um aravap poruḷa “kampalai, cummai, kali & aḻuṅkal, those four have ‘noise, clamour’ as value”. 24 TC350c: avaṟṟu uḷ / aḻuṅkal irakkamuṅ kēṭum ākum “Among them, aḻuṅkal is ‘pity’ or ‘loss’ ”. 25 TC352c: ceḻumai vaḷaṉuṅ koḻuppum ākum “ceḻumai is ‘fertility’ and ‘fat’ ”. Acta Orient. Hung. 63, 2010



7. Citation Form and Inflection of uric cols Although space and time do not allow me to deal at length with this question, a few explanations at least must be given concerning the morphology of those 120 items which the TC labels as uric cols. We can see, inside 10, 11 and 12, that the entities U32 (kampalai), U52 (cummai), U35 (kali), U5 (aḻuṅkal), U54 (ceḻumai) which Cēṉāvaraiyar comments upon [mentioned here in the order of occurrence] are represented in the citations which he provides by linguistic strings (here in boldface) inside the following phrases: kampalai mūtūr (10), cummaiyoṭu (10), kali koḷ āya(m) (10), aḻuṅkal ūr (10), aḻuṅkiṉaṉ allaṉ ō (11), kuṇaṉ aḻuṅkak (11), ceḻum pal kuṉṟam (12), ceḻun taṭi (12). It must be remarked that these examples (given by Cēṉāvaraiyar) are probably a perfect illustration for what the TC told us in 3 (section B), namely that the linguistic items called uric col can be “concretely indistinguishable from either noun or verb” (peyariṉum viṉaiyiṉum mey taṭumāṟi). We see indeed that there is every reason to say that – –

cummai looks like a noun (because it has taken the -oṭu case suffix), aḻuṅkal looks like a verb when it is found as aḻuṅkiṉaṉ (a finite form) or as aḻuṅka (one of the types of converb).

However, how to analyse the other cases, where the item appears as the first component in a Noun-phrase? One is tempted to think that this is in fact what is described by 9 (see section 5). If we take for instance “celum pal kuṉṟam” (12) and “celun taṭi” (12), the component in bold face is what TE482i calls a kuṟaiccoṟ kiḷavi “linguistic item (kiḷavi) which is an incomplete word” (cf. supra 9). And TE482i simply tells us that it has to be learnt from usage that the item for which the citation form is ceḻumai becomes ceḻum (or ceḻun) whereas the items aḻuṅkal and kampalai remain unchanged in aḻuṅkal ūr and kampalai mūtūr. Finally, as a conclusion to this section, it must be said that another commentator to the TC, whose name is Teyvaccilaiyār, has put forward the hypothesis that the correct interpretation of uric col is tātu (i.e. skt dhātu) “root” (see Vēṅkaṭācalam 1929/1984, p. 190). As a comment, one might want to say that there must certainly have been many people who regretted that Tamil grammarians never compiled a dhātupāṭha. However, the Uriyiyal, as it was transmitted to us, could never play the role of a dhātupāṭha.

8. Posterity of the Uriyiyal in Later Grammars As appears from statements such as 13, Cēṉāvaraiyar seems to rely for his explanations of language phenomena on an ontology, on a classification of all the things which exist. It is certainly an open question to find out whether in doing that he is faithful to the original Tolkāppiyam perspective. I shall not try, however, to discuss

Acta Orient. Hung. 63, 2010



the philosophical systems involved.26 I am concentrating on the actual lists of words (U-type and P-type), trying, however, to preserve a general perspective. While doing that, a concern which arises naturally is to examine what happened after the age of Tolkāppiyam. Among the later grammars, the 11th-century Vīracōḻiyam does not contain a section devoted to uric col, but we find again a section on uric cols in the Naṉṉūl, an early 13th-century grammar. That section is, however, quite different in character from the one we had in the Tolkāppiyam, as can be seen from its first sūtra, which reads: (14) palavakaip paṇpum pakarpeya rāki / orukuṇam palakuṇan taḻuvip peyarviṉai / oruvā ceyyuṭ kuriyaṉa vuriccol (N441m) “Qualifying words are the names of the various kinds of properties which, signifying one or more attributes (of matter and spirit), are the inseparable adjuncts of Nouns and Verbs, and belong especially to poetry” (Lazarus). The ontological character of this chapter is rather marked, as can be seen from the following sūtra, which reads. (15) uyiruyi rillatām poruṭkuṇam paṇpē (N442m) “By ‘property’ are meant the qualities of both animate and inanimate objects” (Lazarus). And as a matter of fact, the following 12 sūtras are devoted to enumerating the types of living beings (based on the number of senses) and their principal qualities, as well as the qualities of inanimate bodies. Only after that do we find something which has a slight resemblance to the Uriyiyal of the TC, but limited to 3 sūtras. And this is concluded by an encouragement to make use of dictionaries (or rather Thesauri/Nikaṇṭu) such as the Piṅkalam.

9. Embedding of the Uriyiyal inside the Tivākaram In what follows, I shall not, however, examine the Piṅkalam, but rather an older specimen of Tamil kōśa literature, namely the Tivākaram, which is often said to belong to the 8th century A.D. (see James 2000, p. 62). I shall rely for that examination on the 1990–1993 edition, which is what comes closest to a critical edition. This is a work divided in 12 sections, the first 10 being devoted to various categories, such as Gods (chap. 1, with 159 sūtras), Human Beings (chap. 2, with 251 sūtras), etc.27 while


See, however, Chevillard (2008, pp. 444– 462) for a sketch of Cēṉāvaraiyar’s “vision du monde” (world view). 27 The other chapters are: 3. Animals (217 sūt.), 4. Plants (217 sūt.), 5. Locations (180 sūt.), 6. Various Things (114 sūt.), 7. Artificial Objects (206 sūt.), 8. Qualities (213 sūt.), 9. Actions (214 sūt.) and 10. Sounds (130 sūt.). Acta Orient. Hung. 63, 2010



the 11th and 12th stand apart, because they deal respectively with polysemic terms28 and with collections.29 The first remark to be made is that almost all of the 120 U-type terms from the Uriyiyal are found inside the Tivākaram. Among the few which I was not able to find, there is U17 (urum), U74 (nām), U92 (pēm), which all three appear in TC365c,30 U107 (yāṇu) and U102 (murañcal). There are also cases where an item might be considered to appear under a different form, like for instance U83 (payappu) which might be considered to be represented by another equivalent citation form, payattal seen in T1639, among the quasi-synonyms of peṟutal.31 The second remark is that the chapters of the Tivākaram which appear the most often when one searches for one of the 120 U-type terms are, in this order, the 8th chapter devoted to “Qualities” (paṇpu), which is concerned at least 123 times, the 11th chapter (“Polysemic Terms”), at least 55 times, the 9th chapter (“Actions”), at least 37 times, the 10th chapter “Sounds”, at least 18 times.32 The fact that this adds up to more than 120 hits simply means that the Tivākaram has greatly expanded the seed contained in the Uriyiyal and quoted many more possible meanings of those lexemes. The third remark is that if we examine the 100 P-type terms, in terms of their occurrence inside the Tivākaram, we have a similar phenomenon: the 8th chapter is the most often found, but other chapters are also found. This shows that there is a kernel of truth in the statements of the Naṉṉūl quoted earlier (see 14 and 15),33 but that it does not correspond to the whole truth.34 It also appears that, if we examine the distribution of the items corresponding to the 3 parts of the triad, icai kuṟippu paṇpu (“sound” “idea” “quality”), in which kuṟippu is, as already said, by far the most often used (provided we trust Cēṉāvaraiyar), what we find is the following: – –

the very few terms labelled icai “sound” are found, as expected, in the 10th chapter (“Sounds”) the (not many) terms labelled paṇpu “quality” are found, as expected, in the 8th chapter (“Qualities”) 28

Acting as a sort of (potential) index to the preceding chapters, it contains 383 sūtras. Whereas in the first ten chapters, each sūtra contains a (simple) term and its (more sophisticated) quasi-synonyms, and has thus a semantic unity of content, in this 11th chapter, the emphasis is on the semantic ambiguity of (difficult) forms. 29 Among its 234 sūtras, 10 are devoted to pairs, 23 to triads, 25 to groups of four items, etc. 30 However, the 1840 edition of Tivākaram had incorporated the Tolkāppiyam sūtra as an explicit addition, whereas the 1990 – 1993 edition that we use has relegated in the apparatus and removed it from the main text (see p. 491, No. 204*, just after T1484). 31 This would testify of the progress in the definition of a standard for the citation of lexemes. 32 I have also found connections with other chapters: 2nd chapter (“Human Beings”), 9 times; 5th chapter (“Locations”), 8 times; 7th chapter (“Artificial Object”), 6 times. 33 This also explains why some have proposed to translate uric col by “adjective”, although this seems quite inappropriate from the point of view of the TC and although a study of what “adjectives” are in Tamil yields quite different results (see Chevillard 1992). 34 Lexicographical endeavours cannot be completely characterised by using terms borrowed from philosophical systems. Acta Orient. Hung. 63, 2010



the terms labeled kuṟippu “idea” are found massively in the 8th chapter but also in significant numbers in the 9th and marginally in other chapters (including the 10th).

This again shows how any scheme which a practice-oriented discipline (such as is Tamil lexicography, as it started with the Tivākaram) ends up disregarding some of the categories which it had inherited, because they do not appear to be usable (as seems to be the case with kuṟippu, which does not give its name to any chapter in the Tivākaram. The fourth remark is that there is a change of perspective when one goes from the Uriyiyal to the Tivākaram. The main task is no longer to explain “hard words” but to enumerate in one sūtra as many quasi-synonyms as possible for a basic notion. But it often happens that some of the elements in the enumeration are polysemic and are found in several sūtras. The global structure thus created can also be described as a graph.35

10. A Few Conclusions I have here only scratched the surface of a subject, which is important for Tamil philology, as I would now like to emphasise. It may not be realised by every modern day student of Tamil literature that the texts which we have been discussing here, the Tivākaram and the Tolkāppiyam, as well as many others, were memorised by the students of the past.36 In a short auto-biographical account (A Life-Time for the Cause of Tamil, 2009), the late T. V. Gopal Iyer also emphasises that “the Tamil world considered only strong memorising power as the very great strength of Tamil scholars”. Modern philologists are not very likely nowadays to memorise a thesaurus of 9500 words, divided in 2518 sūtras, as is the Tivākaram. It should not, however, deter them from studying it, as well as other thesauri, in order to be able to imagine what it is to live with the literature of a classical language as a part of oneself. They can also try to understand and modellise its inner structure, with the help of modern tools such

35 See Chevillard (forthcoming) “Examining verbal forms inside the Tēvāram, in the light of the vocabulary found inside the 9th chapter of Cēntaṉ Tivākaram”, communication to be presented at the INFITT 2010 Conference (Coimbatore, June 2010). 36 In the preface to a 1968 edition of the Piṅkalam, the anonymous editors whose collective signature is “caivacittānta nūṟpatippuk kaḻakattār” have written what follows: “There was requirement, agreed upon among the teachers of ancient days, that those who wished to study grammatical and literary works and become proficient in them must first learn by heart the thesauri. That scheme was in circulation until a recent past. […] Unlike the dictionaries of today, the thesauri of those days did not have to be consulted from time to time. All those who had learnt the thesauri were moving thesauri” (ilakkaṇa ilakkiya nūlkaḷaik kaṟṟup pulamait tiṟamaṭaiya virumpuvōr mutaṟkaṇ nikaṇṭu nūlkaḷai aiyantiripaṟak kaṟṟu nalla payiṟciyaip peṟṟirukka vēṇṭum eṉṉum kaṭṭāyat tiṭṭam paṇṭaikkālak kaṇakkāyarkaḷiṭattilē iruntu vantatu. it tiṭṭam aṇmaik kālam varaiyil naṭaimuṟaiyil iruntatu. […] innāḷ akarāti pōṉṟu annāḷ nikaṇṭiṉai avvappōtu puraṭṭip pārkka vēṇṭuvatillai. nikaṇṭukaṟṟār aṉaivarum ulavum nikaṇṭarāy vāḻntu vantaṉar).

Acta Orient. Hung. 63, 2010



as those created by Graph theory explorers.37 That in turn might place them in a better position for understanding all the subtle word plays that the gifted scholars of the past were able to insert in their subtle compositions, to the delight of their fellow scholars. This is how a classical culture is transmitted across the ocean of Time. 11. Bibliography Cāminātaiyar, U. Vē. (patippāciriyar) (1946): Naṉṉūl Mūlamum Mayilai Nātaruraiyum. Chennai, Iraṇṭām patippu. Caṇmukam Piḷḷai, Mu. – Cuntaramūrtti, I. (patippāciriyarkaḷ) (1990 & 1993): Tivākaram. 2 Vols. Chennai, Ceṉṉaip Palkalaik Kaḻakam. Chevillard, Jean-Luc (1992): Sur l’adjectif dans la tradition grammaticale tamoule. Histoire Epistémologie Langage (St Denis, Presses Universitaires de Vincennes) tome 14, fasc. 1, pp. 37 – 58. Chevillard, Jean-Luc (1996): Le commentaire de Cēṉāvaraiyar sur le Collatikāram du Tolkāppiyam. Pondicherry, Institut Français de Pondichéry– Ecole Française d’Extrême-Orient (Publication du Département d’Indologie, N° 84-1). Chevillard, Jean-Luc (2004): Ideophones in Tamil: A Historical Perspective on the X-enal Expressives (Olikkuṟippu Āṟṟuppaṭai). In: Chevillard, J.-L. – Wilden, E. (eds): South Indian Horizon: Felicitation Volume for François Gros on the Occasion of his 70th Birthday. Pondicherry, Institut Français de Pondichéry (Publication du département d’Indologie, No. 94), pp. 407 – 433. Chevillard, Jean-Luc (2008): Companion Volume to the Cēṉāvaraiyam on Tamil Morphology and Syntax (Le commentaire de Cēṉāvaraiyar sur le Collatikāram du Tolkāppiyam : Volume 2). Pondicherry, Institut Français de Pondichéry– Ecole Française d’Extrême-Orient (Collection Indologie, 84.2). Gopal Iyer, T. V. [Kōpālaiyar, Ti. Vē.] (patippāciriyar) (2003): Tolkāppiyam Eḻuttatikāram Iḷampūraṇam. Chennai, Tamiḻmaṇ patippakam. Gopal Iyer, T. V. [Kōpālaiyar, Ti. Vē.] (patippāciriyar) (2005): Vīracōḻiyam. Śrīraṅkam, Śrīmat Āṇṭavaṉ Ācciramam. Gopal Iyer, T. V. [Kōpālaiyar, Ti. Vē.] (2009): A Life-Time for the Cause of Tamil. In: Wilden, Eva (ed.): Between Preservation and Recreation: Tamil Traditions of Commentar. Pondicherry, Institut Français de Pondichéry– Ecole Française d’Extrême-Orient (Collection Indologie, 109), pp. 13 – 21. Herdan, Gustav (1964): Quantitative Linguistics. Washington, Butterworths. James, Gregory (2000): Colporuḷ. A History of Tamil Dictionaries. Chennai, Cre-A. Knuth, Donald, E. (1993): The Stanford GraphBase. A Platform for Combinatorial Computing. New York, N.Y. – Reading, Mass., ACM Press – Addison-Wesley Publishing Company.

37 In The Stanford GraphBase, Donald E. Knuth explains (1993, p. 6) how, having built a graph containing the 5757 five-letter words which he was able to find in the English language, he was able to ask his computer to try to find a path connecting two words (with only one letter changed at each step of the path). If the two words were “sword” and “peace”, the computer was able to propose the following path (containing only Biblical words): sword, swore, store, stare, stars, stays, slays, plays, plans, plane, place, peace.

Acta Orient. Hung. 63, 2010



Lazarus, J. ([1878] 1977): Nannul [Orthography & Etymology]. Part Two, English translation, with a foreword by P. Kothandaraman. Madras, The South India Saiva Siddhanta Works Publishing Society. Pālacuntaram, Ca. (1988): Tolkāppiyam ārāyccik kāṇṭikaiyurai, Collatikāram. Tanjore, Tāmarai Veḷiyīṭṭakam. Piṅkalantai eṉṉum Piṅkala Nikaṇṭu, 1968, Kaḻaka Veḷiyīṭu 1315. Chennai, Tirunelvēli Teṉṉintiya Caivacittānta Nūṟpatippuk Kaḻakam. Tāṇṭavarāyamutaliyār (patippāciriyar) (1840): Cēntaṉ Tivākaram. American Mission Press. Vaiyapuri Pillai, S. (general editor) (1924– 1936 & 1938 –1939): Madras Tamil Lexicon. 6 Vols & Supplement. Madras, University of Madras. Vaiyapuri Pillai, S. (1982): History of Tamil Lexicography. In: Vaiyapuri Pillai, S. (general editor) (1924 – 1936 & 1938 – 1939): Madras Tamil Lexicon. 6 Vols & Supplement. Madras, University of Madras, Vol. 1, pp. xxv – l. Vēṅkaṭācalam (R.) (ed.) ([1929] 1984): Teyvaccilaiyār Urai. Tanjore, Cukkila Vaikāci, Karantait Tamiḻc Caṅkap Patippu, Tamil University. Wilden, Eva (2008): Naṟṟiṇai. Critical edition, 3 Vols. Chennai, Ecole Française d’Extrême-Orient – Tamiḻ Maṇ Patippakam. Wilden, Eva (2009): Depictions of Language and Languages in Early Tamil Literature. How Tamil Became Cool and Straight. Histoire Épistémologie Langage (Paris), tome 31, fasc. 2, pp. 117 – 141.

Acta Orient. Hung. 63, 2010

View more...


Copyright © 2017 EDOC Inc.